Optimization is essential in generating a suitable access plan for a query. Once a query is parsed, the query optimizer (or simply, the optimizer) analyzes it and decides on an access plan that computes the result using as few resources as possible. Optimization begins just before execution. If you are using cursors in your application, optimization commences when the cursor is opened.

Unlike many other commercial database systems, SQL Anywhere usually optimizes each statement just before executing it. Because SQL Anywhere performs just-in-time optimization of each statement, the optimizer has access to the values of host and stored procedure variables, which allows for better selectivity estimation analysis. In addition, just-in-time optimization allows the optimizer to adjust its choices based on the statistics saved after previous query executions.

To operate efficiently, SQL Anywhere rewrites your queries into semantically equivalent, but syntactically different, forms. SQL Anywhere performs many different rewrite operations. If you read the access plans, you frequently find that they do not correspond to a literal interpretation of your original statement. For example, to make your SQL statements more efficient, the optimizer tries as much as possible to rewrite subqueries with joins.

To learn more about SQL Anywhere query processing, see the white paper "Query Processing Based on SQL Anywhere 12.0.1 Architecture" at  http://www.sybase.com/detail?id=1096047.